The centrality of learning and memory deficits in schizophrenia has been recently appreciated, but the exact mechanism of these difficulties is poorly understood. Memory impairments in schizophrenia have been attributed to attentional problems, executive control deficits in working memory, rapid forgetting, and problems in semantic encoding or retrieval of declarative memory. Components of working memory (e.g., memory for temporal order, spatial localization, immediate memory span) and declarative memory (i.e., long-term memory for factual information) will be investigated in the proposed research because they have been respectively linked to prefrontal or mesial temporal neural systems which, in turn, are leading candidates as regions of neuropathology in schizophrenia. Existing tests of Temporal Recency (working memory) and Item Recognition (declarative memory) will be modified using word and face stimuli to allow comparison of differential deficit and use as activation tasks during functional neuroimaging. These tasks will first be administered to a normative sample of 100 healthy volunteers, and then given with a standard neuropsychological battery to socio-demographically matched samples of 30 normal controls and 30 patients with schizophrenia to establish the reliability and construct validity of the two tasks, and to examine how differential deficits in working or declarative memory may be related to clinical symptomatology in the patient sample (specific aim one). After task validation, Temporal Recency. Item Recognition, and standard baseline conditions will be administered as activation tasks ("neurobehavioral probes") during 15 Oxygen (15O) Positron Emission Tomography (PET) measures of Cerebral Blood Flow (CBF) to a group of 60 normal controls (30 men, 30 women). Administering these tests as neurobehavioral probes provides a means to link brain function with brain structure, test the hypothesis that working and declarative memory components are mediated by prefrontal and mesial temporal regions respectively, and test if selective regional activation is correlated with better performance (specific aim two). Finally, the same PET CBF studies will be performed with a matched group of 60 patients with schizophrenia (30 men, 30 women) to test hypotheses that impaired task performance on both tasks is related to diffuse left-hemisphere over-activation, and that differential deficit on temporal recency tasks is related to severity of symptoms specific to schizophrenia, and differential deficit on item recognition tasks is correlated with positive symptomatology (specific aim 3). Identifying which components of memory are differentially impaired in schizophrenia, and how these impairments are related to brain function and clinical symptoms has potential clinical benefit of informing cognitive rehabilitation strategies and guiding development of regionally specific neuroleptic medications.